chore: uniform sac network sizes

[limbryan]

This PR is to try to make the Deep RL algorithms more uniform. It allows the SAC, DIAYN and DADS implementation to have separate policy (actor) and critic architectures, as is already the case in the TD3 implementation. This allows more flexibility and is needed sometimes for comparison with QD algorithms which do not necessarily have a critic which needs a bigger size while the actor can be smaller.

This PR changes:

• the hidden_layer_sizes to policy_hidden_layer_size and critic_hidden_layer_size for SAC, DIAYN and DADS

Checks

• a clear description of the PR has been added
• sufficient tests have been written
• relevant section added to the documentation (N/A)
• example notebook added to the repo (N/A)
• clean docstrings and comments have been written (N/A)
• if any issue/observation has been discovered, a new issue has been opened (N/A)

Future improvements

for DADS and DIAYN there is also the discriminator or dynamics architecture to expose. For now it is set to the critic architecture

[codecov-commenter]

Codecov Report

Merging #145 (cad8b1f) into develop (6fa19e7) will increase coverage by 0.01%.
The diff coverage is 100.00%.

📣 This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more

@@             Coverage Diff             @@
##           develop     #145      +/-   ##
===========================================
+ Coverage    92.28%   92.29%   +0.01%     
===========================================
  Files          116      116              
  Lines         6763     6772       +9     
===========================================
+ Hits          6241     6250       +9     
  Misses         522      522              

Impacted Files	Coverage Δ
qdax/baselines/dads.py	97.00% <ø> (ø)
qdax/baselines/diayn.py	93.07% <ø> (ø)
qdax/core/neuroevolution/networks/dads_networks.py	94.02% <ø> (ø)
...dax/core/neuroevolution/networks/diayn_networks.py	100.00% <ø> (ø)
qdax/core/neuroevolution/networks/sac_networks.py	100.00% <ø> (ø)
qdax/baselines/sac.py	94.37% <100.00%> (+0.03%)	⬆️
qdax/baselines/sac_pbt.py	96.49% <100.00%> (+0.03%)	⬆️
tests/baselines_test/dads_smerl_test.py	97.18% <100.00%> (+0.04%)	⬆️
tests/baselines_test/dads_test.py	96.96% <100.00%> (+0.04%)	⬆️
tests/baselines_test/diayn_smerl_test.py	97.01% <100.00%> (+0.04%)	⬆️
... and 4 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[felixchalumeau]

Notebooks might need to be updated as well.

[felixchalumeau]

Looks good 👍